城市公交客流量智能识别系统的研究

发布时间：2018-10-14 19:10

【摘要】：改革开放以来,我国的国民经济发生了翻天覆地的变化,尤其是在城市建设方面,城市交通成为影响大中城市发展的关键因素。如今,随着城市汽车的拥有量日益增多,城市汽车化速度迅速发展,导致城市中出现了许多阻碍城市发展的问题,诸如交通拥堵、能源匮乏、环境污染愈演愈烈以及交通事故频繁发生等。如何提高城市交通资源的利用率成为解决上述城市交通问题的关键。鉴于城市公共交通与小汽车相比具有客运量大、投资相对少、资源占用率小、运营效率高、污染较少、人均占道少等优势,加大城市公共交通发展力度,达到城市交通管理的数字化和智能化的目标,并提高公共交通运营管理效率及社会服务水平成为改善城市交通的必经之路。本文旨在基于多传感器阵列踏板获取公交乘客的客流量数据,分析了公交车乘客上下车脚型变化规律,获取了传感器采集数据的特征,提出了基于脚型轮廓的客流数据判断准则,通过人体运动学原理进行上下车方向识别,并采用BP神经网络算法对原始数据进行预处理和智能识别,计算公交乘客上下车数量。文章对系统的硬件设计和识别软件算法的实现方法作了详尽阐述。最后对算法作了系统测试,准确率达到93%,具有很强的可靠性。
[Abstract]:Since the reform and opening up, the national economy of our country has undergone earth-shaking changes, especially in the aspect of urban construction, urban traffic has become a key factor affecting the development of large and medium-sized cities. Nowadays, with the increasing number of cars in cities and the rapid development of urban motorization, there are many problems that hinder the development of cities, such as traffic congestion and lack of energy. Environmental pollution is becoming more and more serious and traffic accidents occur frequently. How to improve the utilization rate of urban traffic resources becomes the key to solve the above urban traffic problems. In view of the fact that urban public transport has the advantages of large passenger traffic, relatively low investment, low utilization of resources, high operational efficiency, less pollution, less per capita traffic, and so on, the development of urban public transport should be strengthened. To achieve the goal of digitalization and intelligence of urban traffic management, and to improve the efficiency of public transportation management and the level of social service become the only way to improve urban traffic. The purpose of this paper is to obtain the passenger flow data of bus passengers based on multi-sensor array pedal, analyze the changing law of bus passengers' getting on and off feet, and obtain the characteristics of the data collected by the sensors. This paper presents a criterion for judging passenger flow data based on the profile of foot, and uses the principle of human kinematics to recognize the direction of boarding and disembarking, and uses BP neural network algorithm to preprocess and intelligently recognize the original data, and calculates the number of passengers on and off bus. In this paper, the hardware design of the system and the realization of recognition software algorithm are described in detail. Finally, the algorithm is tested systematically, and the accuracy is 93%, which has strong reliability.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.17;TP391.4;TP183

【参考文献】